DanaRabie2002/code-similarity-checker

Fix: wait for tokenizer.json file to appear before loading

Fix: load tokenizer.json directly to avoid tiktoken dependency

Fix: add protobuf to dependencies

Fix: use PreTrainedTokenizerFast to match tokenizer_config

Fix tokenizer registration for modernbert

fix same name error

Register modernbert classes in app.py to fix from_pretrained error

Fix download path to use /tmp for zip files

Disable TorchInductor to fix getpwuid error

Fix filename case for modernbert files

Fix absolute imports for Hugging Face compatibility

Trigger rebuild

Fix all imports in ModernBERT for Hugging Face deployment

Fix relative import for ACT2FN

Add ModernBERT model integration with proper package structure

Update app with ModernBERT support and config

Update app.py to use ModernBert with /tmp model paths

Pin transformers==4.52.4 and simplify Dockerfile

Fix Dockerfile and upgrade transformers

Final version with fixed model loading (local_files_only)

Fix: absolute paths and local_files_only for model loading

Fix: Extract models into /tmp to avoid permission denied

Fix: save zip to /tmp to avoid permission error

Fix: expose port and launch app in Dockerfile

Fix: cleaned up merge conflict in requirements.txt

Add Dockerfile for Hugging Face deployment

file edited

Trigger Hugging Face rebuild

Fix: add Hugging Face metadata block

Force README update with metadata for Hugging Face

Add Hugging Face config metadata to README

Final deployment-ready update for Hugging Face Spaces

Final working version with layout and CSS fixes

Update layout and fix centering

Update app.py to download model from GitHub Releases

Final graduation project version

Auto model transformers' BERT implementation

Fix: switch model download URLs to GitHub release assets

Fix Dropbox direct download links for model zips

Fix import for ModernBert using transformers' BERT implementation

Use dev version of Transformers for ModernBERT

Fix: use ModernBertForSequenceClassification and dynamic port binding

Reduced model size to fit Standard plan

Fix: updated to use correct port for Render

Update app.py with Dropbox model download links

Final updated version with model downloading and cleaned code

Add files via upload

Initial commit